Breaking down IT silos. That seems to be the topic du jour with all the talk of DevOps, Agile, Business Service Reliability and driving customer satisfaction. While the premise is good, particularly driving customer satisfaction, it’s not always easy to break out of individual silos to work as a cohesive team, as culture and other process issues can get in the way.
Making wholesale changes to the way an IT organization works in a short amount of time can be tough to pull off, so why not start with smaller changes? One place to start that can pay immediate dividends is having the application performance management and capacity management teams work together. The group running APM is responsible for ensuring application and business service reliability today and has a great historical view of performance, while the capacity planning team needs reliable ways to forecast infrastructure needs to support growth and drive customer satisfaction in the future. It’s a natural fit.
From a technology perspective, it can work too for those who have APM in-house and a true capacity planning tool, not just a spreadsheet — a recent survey CA Technologies (NASDAQ: CA) conducted with IDG Enterprise revealed some 60 percent of respondents were still using “old-fashioned” spreadsheets for capacity planning. The capacity planning tool can pull historical performance data from the APM system to help build more accurate models and run an unlimited number of what-if scenarios to understand how applications will perform under different real-world conditions.
At CA World 2013 in April, an attendee was talking about how they upgraded a bunch of application servers in the application delivery environment. When the new servers were turned on, the application actually slowed down, despite the added application server horsepower. The problem? Those new applications servers were too good at their jobs and were overwhelming the backend database systems. In this example, APM helped figure out where in the slow application the problem was occurring; having a capacity planning tool in place that could simulate the workload across the entire application delivery network could have helped avoid the overloaded database in the first place.
How can you benefit from the marriage of APM and capacity planning? Five ways:
1. Reduce costs
Who doesn’t like to save money? Unfortunately, IT organizations tend to over-provision an application delivery environment in anticipation of a major spike (e.g. Black Friday, new iPhone sales, when tickets for a big show go on sale) when they may not need to roll out as much hardware or buy as much cloud capacity.
One example: A major telecommunications provider preparing for an iPhone rollout a couple of years ago realized its systems were vastly underused. So instead of buying more hardware, it reduced redundant systems, doubled utilization rates and cut costs by $20 million – all while successfully handling the anticipated spike.
2. Minimize risk
One of the reasons IT over-provisions systems is that it’s an almost guaranteed way for ensuring the system can handle any load. But as noted above, it’s costly. By basing capacity planning on real-world performance data gleaned from the APM system, organizations can more accurately model future capacity needs (be it data center hardware or cloud capacity) that will handle anticipated growth and demand.
3. Improve the quality of applications
In the age of faster, better, cheaper, being able to rollout new applications and services into production that work as expected is essential. Running into unexpected performance issues or having to roll back an update is a killer. Using APM in pre-production can help find performance issues in testing while capacity planning can provide the predictive analytics needed to simulate the real-world impact of the new service or update in production. Knowing how the application is going to work is far better than hoping it will.
4. Continually monitor and right-size your applications
Once rolled into production, the combination of APM and capacity management systems continually works to assure that you deliver the quality experience end users, such as customers and employees, expect. Production environments are anything but static, so being able to assure performance with APM and adjust capacity when needed based on real data rather than rules of thumb is essential.
5. Align IT more closely to the business
Business services and the applications/infrastructure that support them are the lifeblood of the business. IT and the business must be closely aligned to meet corporate goals and continue future growth. By using APM and capacity management in tandem, IT organizations can monitor current performance, find problems quickly and ensure the application delivery environment is right-sized and run cost efficiently to support business operations and meet customer demand.
If you’re looking to make that shift from silo-based IT management to one that’s aligned to drive reliable services and a great customer experience, then marrying APM with capacity management efforts is a great starting point.
Jason Meserve is Product Marketing Manager at CA Technologies.